Agentic AI Engineer
On-siteGurugram, Haryana, India
Job Summary
Design, develop, and deploy production-grade agentic AI applications for enterprise use cases using LangGraph, LangChain, and DeepAgents. Build complex workflows including state machines, conditional routing, memory, and human-in-the-loop systems while implementing RAG pipelines, context management, and tool-use strategies. Develop reusable agent skills, tools, and components with versioning and guardrails, alongside LLM observability and tracing via Langfuse or LangSmith. Implement evaluation datasets, regression testing, and quality monitoring for AI systems. Design and maintain modern data engineering pipelines, including ETL/ELT, batch, and streaming workflows to support data warehouses, lakehouses, and orchestration platforms. Establish data quality frameworks covering validation, lineage, freshness monitoring, and anomaly detection. Design semantic layers, metrics models, ontologies, and knowledge graphs for AI applications. Enforce security and governance practices including access control, auditability, and PII handling in collaboration with Security, Platform, Data, and Compliance teams. Translate ambiguous business requirements into scalable, maintainable, and production-ready AI solutions.
Required Qualifications
- Proven experience building and deploying production-grade Agentic AI / LLM applications
- Strong hands-on experience with LangGraph and LangChain
- Experience with agent orchestration patterns including state machines, conditional routing, memory, tool calling, and human-in-the-loop workflows
- Experience with DeepAgents or similar multi-agent/long-horizon AI frameworks
- Practical experience with Langfuse, LangSmith, or similar LLM observability and evaluation platforms
- Strong understanding of LLM application development, RAG, prompting, context management, tool use, and model selection
- Advanced Python programming and software engineering skills
- Experience with testing, CI/CD, Git, code reviews, and production software development practices
- Experience designing and operating ETL/ELT, batch, or streaming data pipelines
- Understanding of data warehouses, lakehouses, and data orchestration platforms
- Experience implementing data quality, validation, lineage, freshness monitoring, and anomaly detection
- Experience designing semantic layers, ontologies, metrics models, or knowledge graphs for AI applications
- Experience building reusable AI tools, agent skills, guardrails, and interfaces
- Understanding of AI security, governance, access control, auditability, and PII handling
- Strong communication and problem-solving skills with the ability to translate business requirements into technical solutions
- Agentic AI
- Generative AI
- LLM
- LangGraph
- LangChain
- Python
- RAG
- Prompt Engineering
- LLM Orchestration
- AI Agents
- Tool Calling
- Multi-Agent Systems
- Human-in-the-Loop
- Context Management
- Langfuse
- LangSmith
- LLM Evaluation
- AI Observability
- Data Engineering
- ETL/ELT
- Data Pipelines
- Data Quality
- Data Lineage
- Semantic Layer
- Knowledge Graphs
- CI/CD
- Git
- AI Governance
- PII Handling
Desired Qualifications
- Experience with LLM fine-tuning, model distillation, or model serving infrastructure
- Familiarity with AutoGen, CrewAI, LlamaIndex, or other agent frameworks
- Experience with MLOps / LLMOps, including deployment, monitoring, and lifecycle management
- Experience with offline and online AI evaluation, A/B testing, and metric-driven optimization
- Experience working in finance, healthcare, life sciences, or other regulated industries
- Contributions to open-source AI, agentic AI, or data engineering projects
- Experience designing multi-agent architectures for complex enterprise workflows
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.